Recent evidence suggests that linoleic acid (n-6) alone may be insufficient to fully satisfy essential fatty acid (EFA) nutrition during late fetal and early posnatal life. A possible need for omega-3 fatty acids (n-3) is suggested by the accretion rates of n-3 derivatives in the developing fetus, the altered fatty acid pattern and abnormal visual and mental function observed on an Alpha linolenic acid free parenteral fat solution in the human; and evidence of abnormal visual acuity threshold of newborn Rhesus monkeys fed a linolenic acid poor diet. Other animal studies suggest that omega-3 fatty acid deficiency in the perinatal period affects not only brain lipid composition, but various cortical functions including visual acuity and learning capacity. Most milk formula for prematures, as compared to human milk, are lower in oleic acid, higher in linoleic, have little & linolenic acid and virtually no omega-3 fatty acids of greater than 18 C chain length. This study will evaluate potential omega 3 fatty acid deficiency in the VLBW human neonate by studying plasma and membrane fatty acid composition. As functional indicators of deficiency we will measure membrane fluidity and visual response to light stimulation. Four groups of 15 VLBW (less than 1500g AGA) each will be studied from day 10 until day 45 or discharge from the nursery. Group A will receive human milk; group B, artificial formula with a 18:2n6/18:3n3=82; group C, a formula with a n6/n3=12 similar to that in human milk and group D a formula that will include long chain omega-3 fatty acids (20:5n3 and 22:6n3) in the proportion found in human milk. Blood samples and cheek cells will be obtained on days 10 and 30 and at discharge. Fatty acid composition of plasma lipid subclasses, RBC and cheek cell phospholipids will be evaluated using TLC for separation and capillary GLC for fatty acid methyl ester identification. Membrane fluidity will be assessed by incorporating fluorescent probes measuring polarization and lifetime in intact RBC and ghosts. Cholesterol, phosphorus and RBC membrane lipid peroxidation will also be measured. The electroretinogram and visual acuity cortical evoked potentials will be used to evaluate visual function in the four groups at 36 weeks post menstrual age. This study should provide useful information on omega-3 fatty acid requirements of VLBW infants and may improve early neonatal nutritional practices to optimize later mental development and possibly prevent handicaps.